Sleep deprivation is one of the only interventions that have consistently been demonstrated to produce rapid antidepressant effects. The mechanisms by which sleep deprivation brings about rapid antidepressant effects have not been elucidated. It is noteworthy; however, that recent genomic and proteomic studies have shown that acute sleep deprivation rapidly brings about an upregulation of several mediators of neuronal plasticity, most notably CREB and BDNF. Intriguingly, these very same molecules are upregulated by chronic antidepressants, and may underlie the delayed therapeutic effects of most antidepressants. Additional investigation of the regulation of CREB and BDNF by sleep deprivation has revealed that these changes are critically dependent upon the activation of the noradrenergic system. This is particularly noteworthy, since the locus coeruleus noradrenergic projection is quiescent only during rapid eye movement sleep (REM), when the target tissues display their greatest sensitivity; indeed, the temporal dissociation between the firing of the locus coeruleus noradrenergic neurons, and the sensitivity of its postsynaptic targets in the cortex may have considerable relevance for the antidepressant effects of sleep deprivation. In this context, biological rhythms have the capacity to temporally dissociate biochemical processes, and imposing a temporal coincidence on normally dissociated events can have striking and unexpected effects. Thus, it is our hypothesis that activating the normally quiescent noradrenergic system during REM sleep (i.e. when its postsynaptic target system displays its greatest sensitivity) will robustly upregulate CREB and BDNF, thereby bringing about a rapid antidepressant effect. We propose to activate the noradrenergic system during REM sleep by infusing an alpha-2 antagonist, yohimbine. Since it is our hypothesis that activating the noradrenergic system during REM sleep will bring about an antidepressant effect by a similar mechanism as sleep deprivation, we will "enrich" our sample with sleep deprivation responders in this pilot study. Patients, ages 18 to 60 with a diagnosis of major depressive disorder, currently depressed without psychotic features will be recruited into this study. This experimental proof-of-concept study has two Study Phases. Study Phase I consists of total sleep deprivation. Responders to total sleep deprivation who subsequently relapse will enter Study Phase II. Study Phase II is a double-blind crossover administration of either intravenous yohimbine or saline solution during REM sleep. The specific aim of this study is to assess the efficacy of a single dose of intravenous yohimbine hydrochloride (0.125 mg/kg given over 3 minutes) compared with placebo in improving overall depressive symptomatology when administered during REM sleep. Our primary hypothesis is that the intravenous use of an antagonist in patients with major depression during REM sleep will activate the LC and thus increase noradrenergic activity during a time when the locus coeruleus is normally quiescent- namely REM sleep. If the hypothesis that the timing of the activation of the noradrenergic system is crucial in the antidepressant effect of sleep deprivation is correct then an acute antidepressant effect should be observed in patients despite minimal to no disruption of sleep.